INDEX

Explanations

references to objects and entities related to physical locations and items

oai_token-act-pair · gpt-4o-mini Triggered by @bot

domain names, organizations, titles

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 NSCoder

-0.66

دانشنامهٔ

-0.58

djangoproject

-0.55

انجليز

-0.54

 CreateTagHelper

-0.54

 mergeFrom

-0.53

RegressionTest

-0.51

UnknownFields

-0.50

 invokingState

-0.50

下载附件

-0.49

POSITIVE LOGITS

胶

0.35

 MessageBoxIcon

0.33

 Hans

0.32

Lew

0.31

 Gros

0.31

്ട

0.29

 getSource

0.29

HD

0.29

AddArg

0.29

|};

0.28

Activations Density 0.069%

references to objects and entities related to physical locations and items

domain names, organizations, titles

No Comments

No Known Activations